Britain isn’t post-work

Morning all,

Apology in straight away. This was going to be a normal TOTCs, but I got sufficiently annoyed about a BBC headline this week that you’ve got a TOTCs special on… the idea that we’re about to be, or should be, post-work. “AI to hit 40% of jobs and worsen inequality, IMF says” it read. Apparently, all our jobs are about to disappear. Again. Which is odd because I don’t see how there can be any left after all the columns/books written wasted on promising us that the robots were going to have destroyed them all by now anyway.

Inevitably the IMF paper itself is more honest than the headline about the huge uncertainties involved, but its real reason for existing is to give those in Davos a big bold nonsense claim to chat about over canapes – and that’s not a good reason for anything. Plus there is a real price to everyone enjoying the post-work chit chat – it distracts us from making actually existing work better.

The therapist says it’s best to channel irritation, so here we go with some reflections from myself and Hannah Slaughter (you’re already following her for labour market insights galore right?) on these never ending end of work claims. The goal? To put what we know is happening to humans, not false certainty about the impact of algorithms, centre-stage.

Happy reading,

Torsten and Hannah

Luxury communism or robot dystopia

We’re all used to post-work predictions. Silicon Valley tech bros reckon the future involves a few engineering geniuses working, while the rest of us chillax on the proceeds of our Universal Basic Income. Leftier tech optimists are as perky but want the machines collectively owned. Communism with a robotic face is the offer, where apparently “luxury will pervade everything as society based on waged work becomes as much a relic as the feudal peasant.” At the less perky end of the spectrum we have predictions of the rise of the job-killing robots.  A decade ago,  one of many widely publicised studies made waves by claiming that 35 per cent of British jobs were at risk from computerisation, or “Robots have taken more than 60,000 jobs from British workers with 15 MILLION more to go” as it was reported back then. Now of course there are specific sectors and occupations that have seen big change. The last bout of significant deindustrialisation, for example, witnessed big reductions in employment in British clothing manufacturing through the late 90s and early 2000s (as an aside, note how much of the change happens via fewer young people entering the sector rather than the stereotype of older workers being laid off).

But just because one sector or occupation shrinks doesn’t mean we’re on the road to a post-work future – it’s not that long ago that our ancestors were all in the farming game. And, anyway, we’ve actually seen less of this kind of drastic sectoral shrinkage since the big robot scare got going. What happened in the years immediately after the 2013 panic, while policy makers held never ending seminars on the robot risks? Employment levels rose year after year, repeatedly hitting record highs.

Being wrong is forgivable, but being distracted far less so. Panicking about a flood of robots was, and is, mad when our actual problem is that the UK invests too little in almost everything – robots included. We install robots at less than a tenth the rate seen in Germany. Plus the years when all the attention was on robots saw large increases in insecure or atypical work that we’ve done almost nothing about.

Today newspaper columns and books repeat exactly the same stories, just about artificial intelligence rather than robots. The coverage of this week’s IMF study inevitably ignored the bit where it said that roughly half the jobs that are ‘hit’ would actually benefit from AI. But the bigger problem is the false confidence the headline numbers give when the truth is we really don’t know what impact AI will bring to the labour market. Firms’ take-up of the technology will depend not just on what becomes technically possible but also on cost, culture, and regulations. All of these are uncertain. Of course, we should keep researching and pondering AI’s impact, but enjoying that pondering isn’t an excuse for ignoring the volume of work that non-artificial intelligence – or people as we used to call them – are doing in the here and now.

Where is the 15-hour week?

One thing that we can be more certain about is that technological progress can reduce the amount of work we do. We have almost two centuries of data to prove this. Even if the advance of technology doesn’t end work in the way the futurologists keep predicting, it has historically, and thankfully, reduced how much of it we do. Back in the 1850, workers on average put in almost 65 hours a week. Today we work roughly half that. As productivity has risen, we have taken some of that improvement in the form of working less rather than earning more (though preferences vary, especially if you’re French or American). This has come in different forms over time. In the 19^th Century the focus was on shorter work days, and then freeing Saturday afternoon for leisure (partially in exchange for workers actually turning up on Mondays, rather than continuing the practice of widespread ‘Saint Monday’ absenteeism). The full two-day weekend didn’t get going until the 1930s, while the priority in the 1960s and 1990s was more paid holidays.

This progress has repeatedly driven predictions that work will cease to be an important part of our lives. In 1930, the great John Maynard Keynes wrote that his generations’ grandchildren could look forward to working “three-hour shifts or a fifteen-hour week”, as scientific progress plus decades of saving and investing would mean all of our economic needs, even if not all our wants, could be easily satisfied. He was right about the direction of travel, but out by 100 per cent on the length of the working week as we approach the 100-year anniversary of his essay. The spread of the 48-hour week after the Great War greatly reduced average working hours, but progress has slowed considerably since then. Average hours worked have stopped falling altogether since the financial crisis, when productivity growth – the ultimate driver of rising pay and falling hours – disappeared.

As a result, our hours worked averaging 32 in every single year for the past two decades (lockdowns aside). Feeling poor has meant we work more, or at least don’t work less. Compare that to the 1950s and 1960s when we were knocking 15 minutes off the working week each year.

We’re actually working more not less

The chart above is usually taken to mean that while average hours have stopped falling, we are still working less than previous decades. But that’s not true. Why? Because the chart leaves something rather important out of the story: women. Fifty years ago, 55 per cent of women worked. Today that has risen to over 72 per cent. Because women generally work fewer hours than men this increase has pushed down the average hours worked, despite it being a huge story about us working far more not less. The total hours worked by an average couple is actually up 15 per cent over the past four decades, and has risen again post-financial crisis.

Too much work, or too little?

Maybe we shouldn’t be surprised post-work discussions are so popular, even if not well grounded. Because, as our last chart makes clear, it reflects the desires of higher earners. The highest-paid tenth of the workforce are more than twice as likely as the lowest-paid tenth to be over-employed – that is, want to work less, even if it comes with lower pay. By contrast, low earners are far more likely to be under-employed and want to work more hours to boost their weekly wages. And this doesn’t capture the many more low earners who either can’t up their hours because of their circumstances (kids or health conditions, say), or choose to do the least work possible to get by because their job is so rubbish. The concentration of hours worked amongst higher earners is relatively new – the longest hours used to be done by the lowest earning men.

An increasing share of young men without kids are working part-time. Is this a sign of post-consumerism or dangerous discontent? And why has it got no air time, compared to wild forecasts about robots? Equally under-discussed is that this concentration of hours worked among higher earners is relatively new – the longest hours used to be done by the lowest earners. This new ‘hours inequality’ is holding back the living standards of low-to-middle income Britain.

Most of you reading this will be at or near the end of your working week. And work is usually off the table conversation-wise at weekends. Fair enough. But in general we need a bit less futurology and a bit more work on… actual work. It’s time to focus on the world we’re living in, rather than one it’s fun to imagine. Because there’s nothing post-work about Britain.